Skip to contents

Profile Likelihood for Beta Coefficients

Source: R/plot_uncounted.R
profile_beta.Rd

Evaluates the log-likelihood and population size over a grid of values for a chosen beta coefficient. When reoptimize = FALSE, the population size \(\hat\xi\) is constant because beta does not enter \(\xi = \sum N^{\alpha}\) directly.

Usage

profile_beta(
  object,
  coef_index = 1,
  grid = NULL,
  reoptimize = FALSE,
  plot = TRUE,
  ...
)

Arguments

object

An "uncounted" object.

coef_index

Integer: which alpha coefficient to profile (1 = intercept).

grid

Numeric vector of values to evaluate. If NULL (default), auto-generates 30 points spanning \(\pm 3\) SE around the MLE.

reoptimize

Logical. If TRUE, optimize all other parameters at each grid point (true profile likelihood). If FALSE (default), hold everything else at the MLE (concentrated profile, faster).

plot

Logical; produce the plot? Default TRUE.

...

Additional arguments passed to plot().

Value

Invisibly, a data frame with columns: value, xi, loglik.

See also

profile_alpha, profile_gamma, profile.uncounted